Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: [1]

Author Topic: A taxonomy of strategies (draft)  (Read 3832 times)

0 Members and 1 Guest are viewing this topic.

HiveMindEmulator

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2222
  • Respect: +2118
    • View Profile
A taxonomy of strategies (draft)
« on: October 17, 2012, 07:27:59 pm »
+6

Preface:

This article is based on the article on Deck Archetypes. I semi-promised that I would have post this a couple weeks ago, but I was busy with real life stuff, and this actually took a really long time. This is probably the messiest article I’ve written, but hopefully even in this state it’s worth something, and it may be easier to fill in all the missing stuff with some comments from the community. I'm going with the increasingly popular "(draft)" tag in the title as an excuse for the mess...

Classifying all types of decks into a small number of archetypes is a difficult task, similar to coming up with a taxonomic sorting of animals. Apparently, making the first distinction between vertebrate/invertebrate makes the most sense in terms of current beliefs about evolution, but then you end up with things like flies and birds, which both fly, appearing very far apart from one another, despite having some very similar characteristics. Since I have no carbon-dated fossil evidence or knowledge of migration patterns, I can’t claim that my way of sorting out these strategies is better than any other way, but it is related to how I think about things, so I hope it’s at least comprehensive, somewhat sensible, and maybe even moderately useful.

Much of the terminology I use is also probably not going to be ideal, so I welcome suggestions for better terminology to use. I’d also appreciate it if you can point out types of decks that I may have forgotten that don’t seem to fit into this taxonomy, though I’m not terribly interested in adding bad "strategies" in, so I'm not concerned if the "village idiot" doesn't fit well. There are too many ways you can do things wrong that it’s not really worth trying to classify all of them, at least for the purposes of this article.

I’m also aware that this is a massive amount of text, so I put a little tl;dr taxonomy list at the bottom so you can see the big picture and only read the sections that interest you. While I’m at it, I’d also like to make a little appeal to the admins to implement collapse/expand type spoilers so that the reader can just collapse and expand whole sections here and minimize the scrolling around you have to do to find the next section.

The Phyla:

I think most people agree about the general idea of the first division in our taxonomy. There’s “engines” (which doesn’t necessarily mean the same thing to everyone), and “things that aren’t engines”, which is the complementary set. I’m going to call this set “rush” strategies. I know people also have differing opinions about this term, but I feel it’s appropriate here, and this is how I will define the two phyla:
A “rush” strategy is any strategy in which you would buy a pure victory card on the third or fourth time through the deck, given the appropriate opportunity. An “engine” strategy is anything else. Typically engines cycle fast enough that the fourth shuffle happens pretty early on, and in the time before that, you have a lot of pieces to collect. Why fourth shuffle? I don’t know. From experience, it seems like a reasonable threshold. Despite being a “taxonomy”, this article isn’t incredibly scientific...

I. Rush strategies

The general idea of a rush strategy is to get a minimal number of pieces into place so that you can somewhat consistently acquire the victory cards of your choice (usually Provinces), and then start buying them quickly. You win by doing one of the following:

1. Winning the VP card split against a similar rush strategy
2. Gaining half the available points before a slower strategy kicks in
3. Getting a lead and then ending the game on piles

Some people may want to split the classes based on the choice of end-game, but I don’t think this really makes sense since generally you don’t want to make the end-game decisions when you initially pick your strategy. The choice of these end-game options will depend on what your opponent decides to do. For instance, if you’re going for a Gardens strategy and it turns into a mirror, you want to win by condition 1, but if your opponent goes Provinces instead, you want to go for 3, in which case you may want to build up a little more. While this leads to 2 differing strategies, I feel they belong in the same class, since the “evolutionary split” doesn’t happen until the game is already underway.

Instead I choose to split rush strategies into 3 different classes, based on the general kinds of cards you're looking to put into your deck:

A. Terminal draw big money
B. Non-drawing terminal big money
C. Kingdom VP

In the following subsections, I will attempt to explain the general notions of these strategies, give some examples, and give a short list of what to look for/avoid.

IA. Terminal draw big money.

I start here because this is an easy one. For a lot of people, this is one of the first types of “strategy” they actually learn. The beauty is in the simplicity. The general idea is to collect 1-3 actions which draw cards, and fill the rest of your deck with treasures. When you have a large hand of mostly treasures, you can very often buy a Province or at least a Gold or Duchy. I have already written about these strategies in fairly heavy detail here.

Needs:
 - Terminal draw
 - Silver/Gold
Likes:
 - Kingdom Treasures
 - Estate trashing with additional benefits (Island, Masquerade, Jack)
 - End-game accelerators (Salvager, Apprentice, Remodel, Trade Route)
Dislikes:
 - Heavy trashing
 - Copper trashing
 - Things that lengthen the game (attacks, kingdom VP cards)

IB. Non-drawing terminal big money

This type of strategy is also typically called a “big money” strategy as it also usually relies heavily on using Silver and Gold to buy Provinces. The key difference from the terminal draw deck is that you can collect many more actions. Usually you don’t want to go over 3-5 terminals, but since you don’t draw a lot of cards dead, you can also add in a bunch of non-terminals. Generally, these strategies are slower at Province collection than terminal draw big money strategies, but they make up for it by having more powerful abilities on the actions. The most typical examples of these strategies involve powerful attack which make it difficult to go for engines -- things like Mountebank and Swindler. It works with lesser attacks as well, depending on how much defense and engine support there is. Often this type of strategy is more of a last resort when you really can’t do anything elegant. It’s probably like the way you first started playing the game. Just collect a bunch of “good” actions, and try to play one every turn. Sometimes there’s simply no way to build an engine, and your best bet is just to mass the strongest cards you can get your hands on, which, in addition to attacks, could be things like Monument or Merchant Ship, or even things like Tribute or Harvest is there is really nothing better going on in the kingdom.

Needs:
 - Good terminals
 - Silver/Gold
Likes:
 - Attacks
 - End-game accelerators
 - Fishing/Walled Village (allowing for extra terminals)
 - Non-terminal draw
 - Light sifting
Dislikes:
 - Heavy trashing
 - Copper trashing
 - Engine enablers

IC. Kingdom VP

The concept here is to accumulate a large number of cheap VP cards instead of saving up for Provinces. Then either end the game on piles, or pump the value of these cards up to the point that you can still outscore a Province player. One of the most well-known examples is the Workshop/Gardens rush discussed in the Gardens article. There are also various Duke strategies which are treated well in the Duke article. There is also often a way to pull it off with Silk Road, especially in the presence of other cheap VP cards like Island, Great Hall, or Tunnel.

Needs:
 - kingdom VP cards
 - ways of gaining multiple cards -- to collect VP cards faster, empty piles faster, to help acquire additional economy to counteract early greening, and/or to collect cards to increase the value of variable-value VP cards
Likes:
 - cards that like VP cards (Crossroads, Baron)
 - Terminal Silvers -- since your deck usually bloats, you can stand more terminals, and usually what you want is either gain or money with any additional benefit you can get
Dislikes:
 - Trashing
 - Card draw (since the average value of the cards you draw will typically be low)



II. Engine strategies

This is where the real fun starts. Generally the appeal of Dominion is playing combos of cards. In the rush strategies, you just go after strong cards perhaps with some synergy with one another, but you rarely get to do something “cool” on your turn. No one posts game logs of themselves going Courtyard big money...

The idea with engines is that you delay the instant gratification of getting an early VP lead in order to put together a strong combination of cards that can either allow for more consistent VP collection, or huge influxes of VPs per turn. It’s important to make the distinction between these two choices when you go about building your deck, because typically the hardest part is figuring out the right time to go green. This decision, like the end-game decisions for rush decks, depends in part on what your opponent is doing, but it often depends even more heavily on the class of engine you are building. I have decided to split the engine phylum into 4 classes:

A. Cycling
B. Mass drawing
C. Empty-draw deck
D. Mega-turn

This choice of class splitting isn’t nearly as obvious as in the rush phylum, but after some thought, I find it the most natural one for me, though I’m sure there’ll be plenty of disagreement here. The general distinction between cycling and mass-drawing is that cycling decks never actually get very large hand sizes, but achieve their goal through at least passing through most of the cards. This is a little iffy because there’s no clear distinction for what is meant by “very large” -- maybe 8-10 cards? Sometimes you mix in a couple Warehouses to cycle, but still draw up most of your deck, and I’d say this qualifies as mass drawing rather than cycling, but it may not be tremendously different from a cycling deck. Additionally, mega-turn decks often look a lot like mass draw decks that just end with a final turn cashing in. In some cases you might feel like you’re mostly building a mass draw deck and just have an opportunity to make it go mega-turn. On the other hand, there are some strategies that are clearly mega-turn that never really draw large hands until the final turn.

In the following subsections, I will attempt to explain the general notions of these strategies, give some examples, and give a short list of what to look for/avoid.

IIA. Cycling

Cycling decks achieve shuffling every 1-2 turns so that you can play each of your cards with much greater frequency than in a rush deck, but they do not draw your entire deck into your hand. I have decided to split cycling strategies into 2 sub-classes: sifting and virtual money cycling. The main difference is that sifting decks cycle with the goal of playing a small number of key cards very often, while virtual money cycling decks cycle with the goal of playing a lot of cards for coins. I will discuss each of these sub-classes separately.

1. Sifting

The idea behind sifting decks is that rather than having a high density of key cards or the ability to play a lot of cards, the goal is to be able to play a small number of key cards a disproportionate amount of time relative to the weaker cards. This is accomplished by using cards that let you seek out other cards while usually discarding others.

Perhaps the most popular example is the Hunting Party deck, discussed here. In these decks, you just get 1-2 copies of a strong terminal, usually 1 Gold, and a bunch of Hunting Parties. Then use the Hunting Parties to play the key cards every turn. Stables can also be used in place of Hunting Party, though it’s much less reliable.

It’s also common to just combine some source of card draw with a separate source of sifting. It’s common to do this with Menagerie and a discarder like Hamlet or Oasis, using the discarder to activate Menagerie so you can draw and play your 1-2 Golds nearly every turn. Apothecary also works great as card-draw in this type of deck. You can find some good examples in MMM's Apothecary article.

It can also be done with a source of terminal draw, particularly, draw-up-to-X like Library. An example of this would be Library/Cellar, where you just get 1-2 Golds and Libraries and then repeatedly play Cellars until you get a hand with both the Library and Gold, and use the Library to fill in the rest of the hand, usually getting you up to $8. Draw-to-X is generally valued because the non-drawing sifting cards tend to decrease hand size. But you can also pull of the same kind of strategy using terminal draw cards that draw a lot of cards, like Smithy, Torturer, or Hunting Grounds.

You can also pull of some cycling strategies with no hand-size increasing whatsoever. An example is Warehouse/Treasure Map, where you use the Warehouses first to line up the TMs, then the Golds. Warehouse/Tunnel naturally works in the same way.

Needs:
 - +Cards --usually something that has the potential to look through at least 3 cards, even if it has to discard some (Warehouse, Cellar, Storeroom, Stables, Hunting Party, Library, Catacombs)
 - Usually discard-for-benefit
 - Villages if the draw or discard cards are terminal
Likes:
 - Cards that seek specific cards/types (Hunting Party, Library, Golem, Scavenger, Scheme, Herbalist)
 - Deck-top sifters (Apothecary, Cartographer, Spy)
 - Draw-to-X if using hand-size decreasing sifters
 - Attacks -- your sifting should allow you to play them every turn, making even weak-seeming attacks like Fortune Teller actually quite potent.

2. Virtual money cycling

Virtual money cycling differs from sifting in that the focus is not on repeatedly playing 1-2 key cards, but rather or playing a lot of cards which each provide money. At times this can allow these decks to produce multi-Province turns.

There are so many variations of this type of deck that after writing this section, I decided I needed to further subdivide into sub-sub-classes (orders? sub-orders? families?) I’m not sure how much I want to push this biological taxonomy terminology...). Anyway, I’m going to futher subdivide based on the primary method of cycling. There are of course some common elements to any of these decks. The most obvious of these is “virtual money” cards, i.e. actions that generate coin. But additionally, to make this kind of strategy work, you will also almost always need trashing or buy/gain, usually both. Otherwise it’s not really possible to get a high enough density of these virtual money cards in time to beat a rush strategy.

I will now briefly discuss the different methods of cycling.

2a. Draw-to-X

The most popular example of a virtual cylcing deck is a Minion deck, where Minions are used to draw up 4 cards at a time and all coin-providing cards are played before playing another Minion to draw 4 new cards. Naturally, this also works with any other draw-to-X card in place of Minion, provided that there is some other way of decreasing hand-size, usually while also providing coin, like Oasis, Fishing Village, or Lighthouse.

Aside: It should be noted that engines relying on draw-to-X cards can be either sifting or virtual money cycling, or maybe even a hybrid. It’s regrettable that these are not in the same sub-class, but I still think it’s the more logical sub-class distinction to make than splitting up based on the specific type of drawing card. After all, then one might complain that Library big money is too far from other Library decks in this taxonomy. It’s simply not practical to sort by the key card, and makes more sense to sort by the goals you have for the deck. In draw-to-X sifting decks, you’re looking much more at being able to draw more cards to find that key card, so Hamlet is more useful than Fishing Village because it draws a card and then allows the draw-to-X card to draw 2 more via its discard. Fishing Village only allows for drawing one extra card, but provides money that you’re more likely to want in the virtual money cycling deck.

2b. Lab Variants

Virtual money cycling strategies can also be done with Menagerie when you have discarding cards, and can even work with other Lab-types. However, non-menagerie Lab-types generally provide less card draw potential than draw-to-X cards in this type of deck, so with those you want to rely more on non-hand-size-decreasing sources of virtual money like peddler-types, particularly Conspirator.

2c. +Cards/+Actions

You can combine villages with terminal draw to get the net effect of having Labs, but this can be a little more difficult to pull off since you end up needing a lot of parts. Generally if its possible to get enough of these cards to make the deck reliable, you can probably get into the mass drawing type of deck.

2d. Single-card cycling (peddler variants)

With sufficient trashing and +buy/gain, you can also sometimes make a virtual money cycling deck work with no hand-size-increasing cards whatsoever! If your deck consists primarily of peddler-types (usually Market/Grand Maket/Highway/Peddler), you can often play enough to be able to buy a Province before your hand become full of victory cards.

2e. Tactician

One special family of strategies is what I call “perpetual Tactician strategies”. (Digression: The term “double Tactician” is commonly used, I guess because you have 2 Tacticians in play every turn, but I feel that term is a little confusing. It makes it sound like you have 2 Tacticians in your deck -- like "double Jack", but in some situations, you may need more, and sometimes you can have 2 Tacticians but still not be going for this type of strategy. I think “perpetual” captures the idea more accurately.) The idea of a perpetual Tactician deck is to start your turn with 10 cards, play some actions that produce $8 (or $11 or $16, or whatever), then play another Tactician to discard the remaining cards to set up the next turn. This requires some trashing and/or cycling to able to consistently play a Tactician every turn, but is very powerful when it gets going.

A recap of the what virtual money cycling decks are looking for:

Needs:
 - Coin-giving actions
 - Trashing -- Coppers and Estates really get in the way of spamming out a lot of actions
 - +Buy/gain -- both to collect all the virtual money and to spend it all
Likes:
 - Draw-to-X cards (Library, Watchtower, Minion), preferably with hand-size decreasing cards
 - Non-terminal draw cards (Menagerie, Laboratory, Shanty Town, Stables)
 - Tactician
 - Attacks -- to prolong the early game since you’re trying to mass cheap cards
Dislikes:
 - Silver, beyond maybe 1-2 to get started

IIB. Mass drawing

Sometimes you can do better than just cycling through your cards. Given the right setup, you can actually draw a large portion of them -- or even all of them -- into your hand. This generally demands a lot from the kingdom. You need some combination of heavy trashing to remove the relatively useless starting cards, and heavy drawing to draw all the other cards that you acquire. Additionally, since you want to fill up this deck with useful cards, you need some way of rapidly collecting cards and/or prolonging the game while you build up. The big things you’re looking for are:
 - Buy/gain
 - Attacks
 - Trashing
 - Kingdom victory cards
If you don’t have at least 2 of these, you’re usually not going to be able to pull off a mass draw deck that is better than just a terminal draw big money or simple cycling deck. The only major exception that comes to mind is Torturer engines. The attack is really strong, and also has a kind of victory point effect built in, and it’s all wrapped up in the draw card as well. There are also some exceptions with Wharf (Wharf+Fishing Village, Wharf+Crossroads), but these too are exceptions. You generally don't want to just think that having a good drawing card can make a mass draw engine.

In addition to the aforementioned enablers, you also need an actual way of drawing you deck. This can be done in one of two ways:

1. Non-terminal draw (lab-types)
2. Terminal draw + villages

Building with non-terminal draw is simpler, since you don’t have to worry about card balance, but then you generally need to be collecting a lot of expensive cards (unless you are primarily drawing off of Menageries or Caravan). With terminal draw+villages, you usually have one of your extra benefits -- like your buy or attack, or even you VPs (Nobles) -- built in to one of the drawing components. Plus, one of these components (usually the village) is cheap enough that it’s possible to collect a lot of pieces more rapidly than in the non-terminal draw engines.

Needs:
 - +Cards
 - Trashing (almost always)
 - +Buy/gain (almost always)
Likes:
 - Attacks
 - Kingdom VPs
 - Villages
Dislikes:
 - Silver, beyond maybe 1-2 to get started
 - Starting Coppers/Estates

IIC. Empty draw pile

There is an extreme way to draw your entire deck every turn even without a lot of +Cards. You can just make your deck extremely small! If your deck is 5 cards or less (excluding in-play durations and cards on mats), you can just keep your draw pile empty, and you don’t even have to physically shuffle! The problem with with these strategies is you can’t really collect victory cards, since unless you can immediately put them on a mat (Island or Native Village), they will screw up your empty draw pile.

The most common example of an empty draw pile strategy is the “Golden Deck”m which is discussed in detail here. There are many variations, but the general idea is that you get your deck down to a Bishop and 4 other cards, allowing you to trash a card (usually a Province) for points and re-buy a replacement. The standard Golden Deck is [Bishop, Gold, Silver, Silver, Province].

Another popular family of examples is the KC-Masquerade pin [link]. Here the idea is that you can destroy your opponents deck and starve him out by every turn reducing his hand to 3 cards and using a King’s Courted Masquerade with an empty hand to take and trash them all. The canonical KC-Masq pin deck is done with GOons: [KC, KC, Goons, Masquerade] and discussed here. Other variations are discussed here Once you have emptied your opponents deck, he will usually resign, but if you must continue, the simplest continuation is to buy out the Copper pile so he can never rebuild his economy, and then rebuild into any deck that can buy victory cards.

Needs:
 - Payoff card with powerful every-turn function (Bishop, KC’d Masquerade)
 - Trashing
Dislikes:
 - Victory cards
 - Card-giving or Handsize attacks

IID. Mega-turn

Mega-turn strategies are strategies focused around hitting one big turn in which you abruptly end the game while collecting enough VPs to win. Often the build-up will be like a mass draw strategy, but this is not the only way to play a mega-turn strategy. Importantly, mega-turn strategies don’t necessarily need reliability. You just need one mega-turn to win. This allows you to do things like building up a powerful Native Village mat or getting all your points from cheap VP cards like Duchies, Estates, or even Coppers with Goons. One good example is the Native Village/Bridge strategy [link].

In general, these strategies focus around collecting and playing a large number of copies of a card that allows you to gain/buy extra cards. These cards are going to be things like Bridge, Goons, Horn of Plenty (and to a lesser extent other workshop variants), City. You can also do it with any +buy card when combined with price-reducers or cards that stack in powerful ways like Highway, Quarry, Coppersmith, or Bank. You need a mix of collecting these cards, and having the ability to draw and play them all. Native Village or Tactician can help with this, but also the same stuff that helps make mass draw or cycling decks can work.

Needs:
 - spammable +buy/gain
 - +Actions (usually)
 - +Cards
Likes:
 - cards that synergize with +buy (price reduction, Goons...)
 - cards that benefit from large hand sizes (Coppersmith, Bank, Cellar...)
 - cards that let you play other cards multiple times (Throne Room, King’s Court, Procession, Counterfeit)
 - cards that can trash themselves for a bonus (Mining Village, Horn of Plenty)
 - all the stuff other engines like

TL;DR Taxonomy
I. Rush
    A. Terminal draw big money
    B. Non-drawing terminal big money
    C. Kingdom VP
II. Engine
    A. Cycling
        1. Sifting
        2. Virtual money cycling
    B. Mass drawing
    C. Empty draw pile
    D. Mega-turn
« Last Edit: October 17, 2012, 08:00:25 pm by HiveMindEmulator »
Logged

HiveMindEmulator

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2222
  • Respect: +2118
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #1 on: October 17, 2012, 07:28:30 pm »
0

[Buffer in case long post gets longer]
Logged

jomini

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1064
  • Respect: +771
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #2 on: October 17, 2012, 11:44:10 pm »
0

A couple of quick comments:
1. Does Kingdom VP really require multiple card gain to work? Take a simple case - Duke/Scavenger. By the time something simple, like BM/Smithy, can hit 8 provinces you can have a fairly easy shot at 11-12 dukes/duchies which is competitive with 8 provinces. I know that gain helps, a lot, but is it really needed? Likewise, if there are discard attacks out (like Militia), doesn't that make it viable to go Duke or Fairgrounds without +gain?

2. So you list perpetual tactician and draw-to-X (I prefer the term limited draw) as examples of virtual money cycling saying you need trashing and coin giving actions. I can think of counterexamples to both of those. For instance, Secret Chamber/Library needs support (e.g. Haggler, Village), but doesn't need to get rid of the dross, Village -> Secham -> Lib is very good for a province and with some sort of +buy wants to play a lot of those three to generate coin from actions. After all, isn't tactician/vault a virtual money cycling deck? Can't you use HoPs with sifting to get around needing +coin actions? Likewise, can't you use things like Expand instead of +coins? Just curios, currently it reads like you are excluding these, and I was wondering if you had a reason or if it might work better with some different phrasing?

3.  I think "empty pile" is a bit misleading. A lot of decks play very similarly but do have things to draw. For instance, a Masq pin is vastly better if you can throw in a cantrip (even Pearl diver) or three to protect against discard/Masq. Likewise, a Golden Fortress deck (Fortress x5, Bishop x4) plays quite similar to a traditional Golden Deck, but racks up more points and does require a very limited bit of draw. I think these type of ultra-lean decks more center around having only 4 non-cantrips than necessarily having no discard.

4. Somewhere you should have some sort of top deck control deck. The obvious example is Kc/Scheme/X. You really don't care what happens to the discard or draw and you can green indefinitely once the combo fully hits. For example once you play Kc/Kc/Sch/Sch/X/X you have an assured province for any terminal silver (barring discard attacks or top deck mucking). Likewise Kc/Sch/Herbalist can top deck a single P. Stone and green out until it can buy provinces. Also, things like Scavenger/Stash - just about every turn after it hits you can pitch your deck, top deck your Stashes, and then buy a province next turn and Scav/Golem/Counting House - play golem -> Scav (discard deck, top deck golem) -> Chou (pick up copper). Kc/Scav/X can function much like Kc/Sch/X, though it does require multiple copies of each card. Other, non-Kc, shots can include things like using Count to set up the draw for next turn (like top decking Village/Embassy) or Mandarin shenanigans (have a Copper, Silver, Gold, Royals Seal, Hop, buy a province, gain a mandarin). Inn tricks, like University(gain an Inn) -> Poorhouse -> Phou (leave 2 actions unplayed) -> buy a province -> shuffle Uni/Phou/Phou/Action/Action out of the discard can also let you manage your top deck and play like an empty draw pile without actually emptying the draw pile.

They aren't common, but top deck control setups can be very strong and even dominant. I'm not sure where to shoe-horn them into the phyla, but I do think they deserve a mention.

All told this looks like a very nice start.


« Last Edit: October 17, 2012, 11:46:13 pm by jomini »
Logged

HiveMindEmulator

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2222
  • Respect: +2118
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #3 on: October 18, 2012, 12:03:20 am »
0

A couple of quick comments:
1. Does Kingdom VP really require multiple card gain to work? Take a simple case - Duke/Scavenger. By the time something simple, like BM/Smithy, can hit 8 provinces you can have a fairly easy shot at 11-12 dukes/duchies which is competitive with 8 provinces. I know that gain helps, a lot, but is it really needed? Likewise, if there are discard attacks out (like Militia), doesn't that make it viable to go Duke or Fairgrounds without +gain?
You're right about this. By the time I was done at the end, I started putting "(usually)" on all the "needs". It's hard to actually say any one thing is completely necessary, and I end up having a lot of "likes". There is a question about Dukes in non-drawing terminal big money. A lot of times just playing this kind of deck with Dukes instead of Provinces works, as long as the terminal doesn't mind a deck with more green than usual. I either need to stick a "(usually)" in there, or add this alternative somehow...

Quote
2. So you list perpetual tactician and draw-to-X (I prefer the term limited draw) as examples of virtual money cycling saying you need trashing and coin giving actions. I can think of counterexamples to both of those. For instance, Secret Chamber/Library needs support (e.g. Haggler, Village), but doesn't need to get rid of the dross, Village -> Secham -> Lib is very good for a province and with some sort of +buy wants to play a lot of those three to generate coin from actions. After all, isn't tactician/vault a virtual money cycling deck? Can't you use HoPs with sifting to get around needing +coin actions? Likewise, can't you use things like Expand instead of +coins? Just curios, currently it reads like you are excluding these, and I was wondering if you had a reason or if it might work better with some different phrasing?
So I think here you're touching on the difference between "sifting" and "virtual money cycling". They are in the same class because they are in many ways similar. But if you're doing something without trashing, its more likely a sifting deck. You going for one payoff card/combo to show up. Village+Lib+SC is not going to rattle off a bunch of SCs unless you have a lot of villages. More likely you drop one, then Library up some money to buy the Province. This is a sifting deck. Vault/Tact does need trashing or gain or you won't the needed reliability, though it is borderline between the 2 sub-classes since it doesn't actually require you to spam multiple cards for coin.
I don't know how you're planning on using HoP in this kind of deck, so you might need to clarify that. Expand decks are also going to be sifting. You don't play a bunch of Expands. You find your 1-2 and play them often. I don't know if this distinction didn't come off clear enough.

Quote
3.  I think "empty pile" is a bit misleading. A lot of decks play very similarly but do have things to draw. For instance, a Masq pin is vastly better if you can throw in a cantrip (even Pearl diver) or three to protect against discard/Masq. Likewise, a Golden Fortress deck (Fortress x5, Bishop x4) plays quite similar to a traditional Golden Deck, but racks up more points and does require a very limited bit of draw. I think these type of ultra-lean decks more center around having only 4 non-cantrips than necessarily having no discard.

4. Somewhere you should have some sort of top deck control deck. The obvious example is Kc/Scheme/X. You really don't care what happens to the discard or draw and you can green indefinitely once the combo fully hits. For example once you play Kc/Kc/Sch/Sch/X/X you have an assured province for any terminal silver (barring discard attacks or top deck mucking). Likewise Kc/Sch/Herbalist can top deck a single P. Stone and green out until it can buy provinces. Also, things like Scavenger/Stash - just about every turn after it hits you can pitch your deck, top deck your Stashes, and then buy a province next turn and Scav/Golem/Counting House - play golem -> Scav (discard deck, top deck golem) -> Chou (pick up copper). Kc/Scav/X can function much like Kc/Sch/X, though it does require multiple copies of each card. Other, non-Kc, shots can include things like using Count to set up the draw for next turn (like top decking Village/Embassy) or Mandarin shenanigans (have a Copper, Silver, Gold, Royals Seal, Hop, buy a province, gain a mandarin). Inn tricks, like University(gain an Inn) -> Poorhouse -> Phou (leave 2 actions unplayed) -> buy a province -> shuffle Uni/Phou/Phou/Action/Action out of the discard can also let you manage your top deck and play like an empty draw pile without actually emptying the draw pile.

They aren't common, but top deck control setups can be very strong and even dominant. I'm not sure where to shoe-horn them into the phyla, but I do think they deserve a mention.
You're right about these. I think they all belong in the "empty draw pile" class, which needs a better name to include them.

Quote
All told this looks like a very nice start.
Thanks for the helpful comments!
Logged

jomini

  • Saboteur
  • *****
  • Offline Offline
  • Posts: 1064
  • Respect: +771
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #4 on: October 18, 2012, 01:24:36 am »
0

Quote
So I think here you're touching on the difference between "sifting" and "virtual money cycling". They are in the same class because they are in many ways similar. But if you're doing something without trashing, its more likely a sifting deck. You going for one payoff card/combo to show up. Village+Lib+SC is not going to rattle off a bunch of SCs unless you have a lot of villages. More likely you drop one, then Library up some money to buy the Province. This is a sifting deck. Vault/Tact does need trashing or gain or you won't the needed reliability, though it is borderline between the 2 sub-classes since it doesn't actually require you to spam multiple cards for coin.
I don't know how you're planning on using HoP in this kind of deck, so you might need to clarify that. Expand decks are also going to be sifting. You don't play a bunch of Expands. You find your 1-2 and play them often. I don't know if this distinction didn't come off clear enough.

Well, more what I'm getting at is you can do a good bit of virtual coin without actually trashing down. Something like Lib/Horse Traders/Hamlet can work instead of the Secret Chamber/Lib/Village. The point is, limited draw engines can work quite well with anything that cleans out the dead cards. Yes, you have to have a high enough density of action balance cards to make it work and yes you need more draw than when you have trashing, but it can work. Take another example, Tr/Oasis/Lib/Iw. Here we can use the Iw to gain all the non-Lib components and we can use Tr/Oases for payload, and hand size decrement. Yeah, you need a lot of thrones to make that work, but with Iw, it isn't exactly hard to get them. Things like Tr, Kc, Fishing Village, and efficient action sources can make a lot Lib/discard shenanigans work.

Yeah, these sort of things aren't as strong a nicely thinned engine humming along without 10 cards of dross, but I think there is a bit of lateral gene transfer as it were between sifting and virtual-money cycling. Enough, I wouldn't be so sure that you absolutely need trashing for the latter.

A HoP example would be something like Counterfeit/HoP/Lib and 6 other cards to get in play; the idea being that you can Counterfeit the HoP, gain a province, and gain a new HoP (gaining 2 provinces on the final turn). Here you can certainly thin out the deck with with something like Chapel, it is just using the activated Hop rather than action-cash to get your payload.

Vault/Tac needs two Schemes, some additional draw (e.g. Lab, Vagrant), or some sifting (e.g. scout, cartographer) in order to be reliable. In any event, even an unreliable Vault/Tac setup can be your best move on some boards.

I'm just not that sure that the lines are that bright between virtual money cycling, sifting, and big draw. I get that you want to talk about how the setup is different for each, but there is a good bit of wiggle room between them. I have certainly built engines that use discard for coin mechanisms/limited draw to get back to do it again and that doesn't seem to be what you have in mind for either sifting or virtual coin; I have also used limited draw to get non-coin actions into play to power up HoPs (e.g. Jack/HoP/Upgrade), though this can skirt the edge of being a megaturn strat.
Logged

HiveMindEmulator

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2222
  • Respect: +2118
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #5 on: October 18, 2012, 02:05:40 am »
0

Yeah, so there is no way for it to be perfectly clear. A lot of strats are going to look like they're something between. I would argue that all the examples you're giving are sifting decks. I'll try to clarify my definitions a bit in the next revision, hopefully coming later this week. I think Vault/Tact is a trouble case. I already though about this a lot before. But hopefully the distinction comes out better. The key idea is something like this:
Sifting decks have some major card/combo they are trying to hit every turn. They don't need to play all the cards, and they don't need to buy a lot of money-producing cards, so you can spend a lot of buys on the sifting mechanism.
Virtual money decks need to buy a lot of coin-producing actions. There are definitely situations where you don't need much, if any, trashing. For example, in 2-player, it's pretty common to just go straight Minions with like no support. But often, without trashing, these types of decks are going to have a speed problem. In general, you can substitute cycling/sifting cards for trashing, but the problem here is that these decks tend to need more actual money producing cards, so if the card isn't dual role like Minion, you really need to trash to get the required density.
Logged

PSGarak

  • Young Witch
  • ****
  • Offline Offline
  • Posts: 135
  • Respect: +160
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #6 on: October 18, 2012, 11:30:25 am »
0

Just a disagreement on terminology: I don't like your definitions of "engine" and "rush" in the Phyla section. While most engines do tend to buy VP after more than 3 cycles through their deck, that's not really a design goal, that's a side-effect of cycling faster. And while your definition of Rush does technically make a contrast against an engine, it's a bit misleading. A BM deck can start greening the same turn as an Engine deck, it's just that by that turn the Engine has gone through it's deck more often. Turns, not deck-throughs, are usually the measure of speed for a deck. Basically, what you're taking as definitions I think are side-effects, and not always true. I also think that "Rush" normally means something narrower than "non-Engine."

I think of an Engine deck as something that wants to play multiple Kingdom cards per turn. This usually, but not always, requires actions, draw, and either trashing or sifting.

When I hear the word "Rush" I think 3-pile. A rush isn't just trying to grab VPs, it's also trying to end the game.

Standard BM and BMX is generally not in either of these categories, which would imply a third phylum. And I think that's appropriate. 3-pile/alt VP decks show very little resemblance to BMX decks, and I don't think it's fair to use "Rush" to describe BMX, because all decks are trying to do something fast.
Logged

HiveMindEmulator

  • Mountebank
  • *****
  • Offline Offline
  • Posts: 2222
  • Respect: +2118
    • View Profile
Re: A taxonomy of strategies (draft)
« Reply #7 on: October 18, 2012, 11:38:32 am »
0

Just a disagreement on terminology: I don't like your definitions of "engine" and "rush" in the Phyla section. While most engines do tend to buy VP after more than 3 cycles through their deck, that's not really a design goal, that's a side-effect of cycling faster. And while your definition of Rush does technically make a contrast against an engine, it's a bit misleading. A BM deck can start greening the same turn as an Engine deck, it's just that by that turn the Engine has gone through it's deck more often. Turns, not deck-throughs, are usually the measure of speed for a deck. Basically, what you're taking as definitions I think are side-effects, and not always true. I also think that "Rush" normally means something narrower than "non-Engine."

I think of an Engine deck as something that wants to play multiple Kingdom cards per turn. This usually, but not always, requires actions, draw, and either trashing or sifting.

When I hear the word "Rush" I think 3-pile. A rush isn't just trying to grab VPs, it's also trying to end the game.

Standard BM and BMX is generally not in either of these categories, which would imply a third phylum. And I think that's appropriate. 3-pile/alt VP decks show very little resemblance to BMX decks, and I don't think it's fair to use "Rush" to describe BMX, because all decks are trying to do something fast.

Maybe the better thing to do is change both terms that I use since they are overloaded with other people's meanings. I'm thinking the main difference is greening when you get a chance (without much buildup) vs building up first. It's true that number of shuffles before greening isn't the thing you think about when you pick your strategy, but you (or I, at least) think about "will I want to buy victory cards early or not until I have my 'engine' going?"

I'm not concerned with speed in number of turns because I'm thinking of the deck building process, which I think is much more concerned with shuffles than turns -- it has to do with how often you play your cards. And anyway, if your big money deck doesn't start buying Provinces until the same turn as the engine deck, it has not succeeded.

I also do talk about why 3-pile is not different from general "rush" strategies. It's an end-game choice -- usually one you're aware of when you start your strategy, but not one you decide on. When you start going Gardens, you're thinking of the three pile. Similarly when you start with a University engine. But that doesn't make those strategies similar in the way I'm classifying them. I'm more concerned with how the deck builds and plays than the specific end condition.
« Last Edit: October 18, 2012, 11:43:15 am by HiveMindEmulator »
Logged
Pages: [1]
 

Page created in 0.061 seconds with 21 queries.